Abstract
Hypertrophic cardiomyopathy (HCM) is a common hereditary heart disease, characterized by myocardial hypertrophy, reduced ventricular cavity, and abnormal myocardial fiber structure. Computed tomography (CT) is a commonly used imaging tool in the diagnosis and risk assessment of cardiovascular diseases. This study aims to conduct an exploratory Mendelian randomization (MR) analysis to investigate the genetic association between the propensity to undergo CT (inferred from genetic variations) and the risk of hypertrophic cardiomyopathy (HCM). In this study, a two-sample MR method was employed. Using the genotype data from the MRC-IEU and FinnGen databases, single-nucleotide polymorphisms related to the phenotype "receiving CT scan" were selected. This phenotype was used as a tool variable for the propensity of CT scan. The CT scan propensity based on genetic prediction was regarded as the exposure factor, and HCM was regarded as the outcome. The associations were estimated using inverse-variance weighting, MR-Egger, and weighted median methods. Additionally, multiplicity tests and leave-one-out analysis were conducted to evaluate the robustness of the tool variable and the stability of the estimated values. In the MR analysis, the inverse-variance weighting method indicated a negative correlation (OR = 2.22 × 10-16, P = .013). However, the MR-Egger and weighted median methods yielded non-significant estimates with wide confidence intervals, reflecting differences in the sensitivity of different analysis methods to potential biases and leading to inconsistent results across methods. The funnel plot and leave-one-out analysis suggested that the set of instrumental variables included in this analysis was statistically robust. This exploratory MR analysis, using a two-sample framework, investigated the genetic relationship between the propensity for undergoing CT scans and the risk of hypertrophic cardiomyopathy. The differences in the estimated values obtained from different analytical methods indicate the inherent complexity involved in using broad non-biological exposure indicators in such studies. These findings are at an early stage and emphasize the importance of using more precise phenotypic definitions in future research to further explore this potential association.